AI and the Future of the Audit

For a century, auditors have tested samples, because examining every transaction was impossible. AI and data analytics remove that constraint. Having started my finance career in audit at PwC, I believe the next decade will change the profession more than the last fifty years.

From samples to full populations

Analytics tools can examine every journal entry, invoice and payment, flagging unusual amounts, timing, users and relationships. Auditors can focus their attention on genuine anomalies rather than random selections.

Smarter risk assessment

Machine learning can compare a client's patterns with prior years and peers, highlighting where misstatement risk is highest. Natural language tools can read contracts, board minutes and correspondence to surface commitments and disputes that affect the financial statements.

Continuous assurance

As companies run on connected systems, assurance can move from a once-a-year exercise towards continuous monitoring of controls and transactions, giving boards earlier warning of problems.

What stays human

• Professional scepticism and the courage to challenge management.
• Judgment on estimates, going concern and complex transactions.
• Understanding the business and its incentives.
• Accountability for the opinion.

New risks for auditors

Auditors must understand the AI their clients use in financial reporting, test the controls around it and make sure their own tools are reliable and properly documented. Data security and confidentiality become even more important when full datasets are analysed.

What changes for audited companies

As audit firms adopt data analytics and AI, companies will be asked for more complete data, often directly from their systems. Clean data, consistent charts of accounts and well-documented processes make audits faster and reduce the number of questions. Companies may also see more targeted queries: instead of a general sample, auditors will ask about specific unusual transactions their tools identified.

Audit quality and regulation

Regulators expect firms to understand and control the technology they use. That means testing the accuracy of tools, documenting how they work and making sure auditors can explain conclusions. AI can increase audit quality by examining more data, but it can also create new risks if outputs are accepted without review. Standard setters are updating guidance to address the use of technology in audits.

Career advice for young accountants

• Learn data analysis, including spreadsheets, databases and visualisation tools.
• Understand business processes and IT controls, not only accounting standards.
• Practise explaining technical findings in plain language.
• Develop scepticism: always ask how you know a number is right.
• Pursue professional qualifications; they remain the foundation of trust.

A leader's view: what audit committees should ask

Audit committees oversee the external auditor, so they should understand how technology is changing the audit of their company. Useful questions include: which areas of our audit use data analytics or AI, and what did those tools find? How were the tools validated, and who reviewed their outputs? Did technology change the audit's scope, timing or fees? What data do you need from us, and how is it protected?

Committees should also look inward. An auditor's technology is only as good as the company's data. If the general ledger is cluttered with inconsistent coding, manual adjustments and spreadsheets outside the system, analytics will generate noise rather than insight. Investing in clean data, standard processes and automated controls makes the audit faster and more valuable, and it improves management information at the same time.

Finally, committees should consider internal audit. The same tools that external auditors use can help internal audit monitor controls continuously, test full populations of transactions and focus attention on the highest risks. An internal audit function that combines data skills with business understanding becomes an early warning system for the board, not only a compliance check. In my experience across audit and forensic work, the organisations that find problems early are those that look at all their data regularly, not only a sample once a year.

Advice for young accountants

Learn the standards deeply, learn data skills seriously and keep your scepticism sharp. The auditors of the future will be part accountant, part data analyst and fully professional.

Common questions

Will AI replace auditors?

No. AI will automate testing and analysis, but auditors remain responsible for judgment, scepticism, conclusions and communication with audit committees.

How is AI used in audits?

To analyse entire populations of transactions, identify anomalies, assess risk, read contracts and documents, and support continuous monitoring.